CuriosAI Submission to the EgoExo4D Proficiency Estimation Challenge 2025
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918088786575360 |
|---|---|
| author | Tanoue, Hayato Nishihara, Hiroki Suzuki, Yuma Hori, Takayuki Takushima, Hiroki Manojkumar, Aiswariya Shibata, Yuki Takeda, Mitsuru Beppu, Fumika Hengwei, Zhao Kanda, Yuto Yamaga, Daichi |
| author_facet | Tanoue, Hayato Nishihara, Hiroki Suzuki, Yuma Hori, Takayuki Takushima, Hiroki Manojkumar, Aiswariya Shibata, Yuki Takeda, Mitsuru Beppu, Fumika Hengwei, Zhao Kanda, Yuto Yamaga, Daichi |
| contents | This report presents the CuriosAI team's submission to the EgoExo4D Proficiency Estimation Challenge at CVPR 2025. We propose two methods for multi-view skill assessment: (1) a multi-task learning framework using Sapiens-2B that jointly predicts proficiency and scenario labels (43.6 % accuracy), and (2) a two-stage pipeline combining zero-shot scenario recognition with view-specific VideoMAE classifiers (47.8 % accuracy). The superior performance of the two-stage approach demonstrates the effectiveness of scenario-conditioned modeling for proficiency estimation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_08022 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CuriosAI Submission to the EgoExo4D Proficiency Estimation Challenge 2025 Tanoue, Hayato Nishihara, Hiroki Suzuki, Yuma Hori, Takayuki Takushima, Hiroki Manojkumar, Aiswariya Shibata, Yuki Takeda, Mitsuru Beppu, Fumika Hengwei, Zhao Kanda, Yuto Yamaga, Daichi Computer Vision and Pattern Recognition This report presents the CuriosAI team's submission to the EgoExo4D Proficiency Estimation Challenge at CVPR 2025. We propose two methods for multi-view skill assessment: (1) a multi-task learning framework using Sapiens-2B that jointly predicts proficiency and scenario labels (43.6 % accuracy), and (2) a two-stage pipeline combining zero-shot scenario recognition with view-specific VideoMAE classifiers (47.8 % accuracy). The superior performance of the two-stage approach demonstrates the effectiveness of scenario-conditioned modeling for proficiency estimation. |
| title | CuriosAI Submission to the EgoExo4D Proficiency Estimation Challenge 2025 |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2507.08022 |